Educational and racial differences in a variety of health related outcomes are large and as yet unexplained. Work characteristics, such as stress, recognition, and autonomy, serve as under-explored mechanisms that may connect education and race to wide-ranging aspects of well being such as mental health, family stability, physical health, and mortality. The newly revamped Dictionary of Occupational Titles (now called O*NET), linked to a data source with measures of health-related outcomes, provides an opportunity to evaluate the utility of occupational characteristics for understanding the mechanisms underlying health disparities. As a test case, we propose to investigate the potential for work characteristics to explain educational and racial disparities in marriage and cohabitation by linking data from the O*NET, 2000 Census, and the 1997 National Longitudinal Survey of Youth. First, this project will use these sources to create a data set describing work demands, rewards, and demographic composition that can be linked to the NLSY using 2002 census codes. Second, it will link these data to the NLSY to document racial and educational in occupational characteristics in early adulthood. Third, the project will investigate the association between occupational characteristics and women's union formation and whether attributes of occupations are linked to educational and/or racial disparities in marriage and cohabitation. Finally, after having refined our data set through the specific aims 1-3, we will make data set describing occupational demands, rewards, and demographic composition publicly available through the NLSY. This project will serve as a foundation for a larger study exploring the influence of men's and women's work characteristics on family formation and stability. PUBLIC HEALTH RELEVANCE: Educational and racial differences in a variety of health-related outcomes are large and, so far, mostly unexplained. Work characteristics serve as one under-explored mechanism that may connect education and race to wide-ranging aspects of well being such as mental health, family stability, physical health, and mortality. As a test case, we propose to investigate the potential for work characteristics to explain educational and racial disparities in marriage and cohabitation by linking data from the O*NET, 2000 Census, and the 1997 National Longitudinal Survey of Youth. Marriage and cohabitation are good outcomes to test given that there is a large literature theorizing a strong influence of paid work on marriage and family life and marriage and cohabitation are linked to other health outcomes ranging from birth weight to adolescent depression to mortality.